Abstract
The purpose of the present study is to exemplify a hypothesis that the in- and out-migration of urban centers is function of the geographic-field, the latter of which is function of both population potentials and urban population within the given group of urban centers. As to the in- and out-migration of urban centers, the variables cover twelve criteria, i. e. gross in- and out-migrants, their respective rates per 1000 population, and growth rates of those migrants for every year from 1970 through 1976 for the nineteen cities, while the variables of both ordinary- and gravity-types of population potentials and urban population for ten coefficients of distance friction from 0.5 to 5.0 with intervals of 0.5, by years and by cities, as to the geographic-field.
The correlation analyses, Pearson's, partial and multiple correlation analyses, show high correlated relationships between the variables of migration and those of the geographic-field as defined above. And, the high correlation (0.9435-0.9991) between these two groups of variables, criteria variables and predictor variables, is also shown by canonical correlation analyses based on their principal components. Moreover, spatial correspondence is found remarkably good between the geographic-field and the in- and out-migration of urban centers, which should imply that spatial order exists within the pattern of that migration, because the geographic-field defined reveals orderly spatial pattern.In that case, the geographic-field operationally defined above accounts for 89.0 to 99.8 per cent of the total variance of the migration.
Thus, it is considered that the proposed hypothesis has added an exemplification to itself.
